Mineral Processing Water Treatment using …Mineral Processing Water Treatment Using!Magnesium Oxide...

6
Mineral Processing Water Treatment Using! Magnesium Oxide Comparative tests show magnesium oxide to be superior to silica sand and gar- net sand for the filtration of several different particulates and to lime for them precipitation of heavy metals. Joseph E. Schiller, Daniel N. Tallman, and Sanaa E. Khalafalla*, Twin Cities Research Center, Bureau of Mines, US. Dept. of the Interior, Minneapolis, Minn. 55817 Removal of Swpeded Solids “This paper describes two applications where magnesium oxide (MgO) can be used to urify process water in two tate dissolved heavy metals. Granular MgO (periclase min- eral) out-performs sand in deep bed filters, since an MgO filter system operates with cycles up to twice as long and gives a cleaner filtrate than a comparable sand filter. To precipitate heavy metals, owdered MgO (magnesia chemical) is used instead o P lime or caustic soda. The MgO-formed sludge cements into a solid mass on standing and occupies only about 20 to 30 pct of the volume as the precipitates produced by more soluble bases.” Media used in deep-bed filters are commonly silica sand and, more recently, a combination of anthracite coal, silica sand, and garnet sand [I, 21. Suspended solids are re- tained in the filter by a combination of mechanisms which include 1) short-range London-van der Waal’s attractive forces between the filtering medium and the suspended particles, 2) chemical bridging caused by flocculation, 3) chemical bonding, and 4) interstitial straining. MgO filters introduce an electrokinetic attractive force which aids particle attachment to the filter grains. Most mineral surfaces are negatively charged in water because their cat- ions have a higher hydration energy than do the anions. Consequently, more cations leave the solid surface, and the surface acquires an excess of negative charges. The surface of MgO is positively charged in water at pH values below 11 [3]. Therefore, an electrokinetic attractive force between the unlike charges assists particle attachment to MgO filters. Negatively charged filter materials can be chemically modified to become positively charged and thereby improve their efficiency. Cationic polyelectro- lytes and alum (KAl(SO,),) have been successfully used to treat diatomite [4, 51. The increased efficiency of a posi- tively charged medium has also been demonstrated for membrane filters [6], and commercial products are avail- able which are based on this concept(Zetapor Membranes manufactured by AMF Cuno Division, Meriden, Corm.).** Granular MgO (periclase) has distinctive properties that make it an excellent filter medium. In addition to its posi- tive electrokinetic charge at neutral pH values, MgO is rel- ways; (1) to filter out suspen B ed solids and (2) to precipi- “Author to whom correspondence should be made. ’* Reference to specific trade names or manufacturers does not imply endorsement by the Bureau of Mines. atively inexpensive and is nontoxic. Its density of3.6 g/cm3 is close to that of garnet sand. Crushed crystalline MgO is irregular in shape, resulting in filter beds with large pore size and high capacity. Removal of Heavy Metals Chemical precipitation of metal hydroxides is the most common way to remove dissolved metals from water. In general practice, sufficient lime or caustic soda is added to raise the pH above 8 or 9, forming a fluffy, suspened hy- droxide precipitate. After adequate settling, the metal sludge is disposed of or reclaimed. Care is required to pre- vent dissolution of the sludge by acidic water or resuspen- sion by agitation. Lime is almost always used because ofits low cost. and coagulants are generally added to improve the settling process. The most common coagulants for hy- droxide precipitates are water-soluble organic polymers. These polymers attach to particles and cause small ones to collect together tci make larger aggregates that settle more rapidly. Chemical precipitation of metals as carbonates with soda ash (sodium carbonate) or sulfides using sodium sulfide are also used where insoluble compounds result [7]. More specific chemicals are sometimes employed for particular metals. Starch zanthate is precipitant with func- tional groups that form several insoluble metal salts [8]. Reverse osmosis, ion exchange, activated-carbon adsorp- tion, cementation, and extraction are other processes ap- plied for metals removal. The solubility, basicity, and surface charge of MgO al- low it to form compact precipitates. Although MgO is quite insoluble in water(the solubility is 5 mdliter), it is a strong base, giving saturated solutions a pH of 10.5. When MgO is added to a solution containing hydrated heavy metal ions, it readily reacts to form metal hydroxides. However, the recipitation reaction apparently occurs at the MgO sur- Face, giving much less occluded water than when a soluble base is used. Usually 5 to 15 minutes are required for MgO to raise the pH, and this gradual increase may lead to larger hydroxide c stals and a more compact precipitate. An- other factor t ri: at may contribute to the compactness of the sludge is the positive surface charge on MgO and the nega- tive surface charge on most heavy metal hydroxides [3]. The resulting mutual attraction ma cause water to be expelled from the spaces between t K e particles givin denser solid. Environmental Pmgreos (Vol. 3, No. 136 May, 1984

Transcript of Mineral Processing Water Treatment using …Mineral Processing Water Treatment Using!Magnesium Oxide...

Mineral Processing Water Treatment Using! Magnesium Oxide Comparative tests show magnesium oxide to be superior to silica sand and gar- net sand for the filtration of several different particulates and to lime for them precipitation of heavy metals.

Joseph E. Schiller, Daniel N. Tallman, and Sanaa E. Khalafalla*, Twin Cities Research Center, Bureau of Mines, US. Dept. of the Interior, Minneapolis, Minn. 55817

Removal of Swpeded Solids

“This paper describes two applications where magnesium oxide (MgO) can be used to urify process water in two

tate dissolved heavy metals. Granular MgO (periclase min- eral) out-performs sand in deep bed filters, since an MgO filter system operates with cycles up to twice as long and gives a cleaner filtrate than a comparable sand filter. To precipitate heavy metals, owdered MgO (magnesia chemical) is used instead o P lime or caustic soda. The MgO-formed sludge cements into a solid mass on standing and occupies only about 20 to 30 pct of the volume as the precipitates produced by more soluble bases.”

Media used in deep-bed filters are commonly silica sand and, more recently, a combination of anthracite coal, silica sand, and garnet sand [ I , 21. Suspended solids are re- tained in the filter by a combination of mechanisms which include 1) short-range London-van der Waal’s attractive forces between the filtering medium and the suspended particles, 2 ) chemical bridging caused by flocculation, 3) chemical bonding, and 4) interstitial straining. MgO filters introduce an electrokinetic attractive force which aids particle attachment to the filter grains. Most mineral surfaces are negatively charged in water because their cat- ions have a higher hydration energy than do the anions. Consequently, more cations leave the solid surface, and the surface acquires an excess of negative charges. The surface of MgO is positively charged in water at pH values below 11 [3]. Therefore, an electrokinetic attractive force between the unlike charges assists particle attachment to MgO filters. Negatively charged filter materials can be chemically modified to become positively charged and thereby improve their efficiency. Cationic polyelectro- lytes and alum (KAl(SO,),) have been successfully used to treat diatomite [4, 51. The increased efficiency of a posi- tively charged medium has also been demonstrated for membrane filters [6], and commercial products are avail- able which are based on this concept(Zetapor Membranes manufactured by A M F Cuno Division, Meriden, Corm.).**

Granular MgO (periclase) has distinctive properties that make it an excellent filter medium. In addition to its posi- tive electrokinetic charge at neutral pH values, MgO is rel-

ways; (1) to filter out suspen B ed solids and (2) to precipi-

“Author to whom correspondence should be made. ’* Reference to specific trade names or manufacturers does not imply endorsement

by the Bureau of Mines.

atively inexpensive and is nontoxic. Its density of3.6 g/cm3 is close to that of garnet sand. Crushed crystalline MgO is irregular in shape, resulting in filter beds with large pore size and high capacity.

Removal of Heavy Meta ls

Chemical precipitation of metal hydroxides is the most common way to remove dissolved metals from water. In general practice, sufficient lime or caustic soda is added to raise the pH above 8 or 9, forming a fluffy, suspened hy- droxide precipitate. After adequate settling, the metal sludge is disposed of or reclaimed. Care is required to pre- vent dissolution of the sludge by acidic water or resuspen- sion by agitation. Lime is almost always used because ofits low cost. and coagulants are generally added to improve the settling process. The most common coagulants for hy- droxide precipitates are water-soluble organic polymers. These polymers attach to particles and cause small ones to collect together tci make larger aggregates that settle more rapidly.

Chemical precipitation of metals as carbonates with soda ash (sodium carbonate) or sulfides using sodium sulfide are also used where insoluble compounds result [7]. More specific chemicals are sometimes employed for particular metals. Starch zanthate is precipitant with func- tional groups that form several insoluble metal salts [8]. Reverse osmosis, ion exchange, activated-carbon adsorp- tion, cementation, and extraction are other processes ap- plied for metals removal.

The solubility, basicity, and surface charge of MgO al- low it to form compact precipitates. Although MgO is quite insoluble in water(the solubility is 5 mdliter), it is a strong base, giving saturated solutions a pH of 10.5. When MgO is added to a solution containing hydrated heavy metal ions, it readily reacts to form metal hydroxides. However, the

recipitation reaction apparently occurs at the MgO sur- Face, giving much less occluded water than when a soluble base is used. Usually 5 to 15 minutes are required for MgO to raise the pH, and this gradual increase may lead to larger hydroxide c stals and a more compact precipitate. An- other factor t ri: at may contribute to the compactness of the sludge is the positive surface charge on MgO and the nega- tive surface charge on most heavy metal hydroxides [3]. The resulting mutual attraction ma cause water to be expelled from the spaces between t K e particles givin denser solid.

Environmental Pmgreos (Vol. 3, No. 136 May, 1984

hours of the run. r allowed use of different materials or

es in series to simulate mixed-media to measure the pressure drop across it possible to determine where pres- taking place. Water containing sus-

es was repared and stirred in a 20-liter

mumin (0.21 to 0.25 m/min). The flocculant and suspen- sion were brought together by a two-channel pump and al- lowed to mix for 8 to 10 minutes before they entered the first filter column. The pressure at the head of each col- umn was monitored using a gauge, and the filtrate turbid- ity was measured every 15 minutes during a run. Filtrate turbidities and pressure increases were essential1 same when a given suspension was filtered throug the same media in either the 2.5- or the 20-cm-diameter filter.

reservoir and pumped t 1: rough the filter at 100 to 120

h the

A

Synthetic suspensions of solids were prepared from finely ground quartz, kaolin, and bentonite. Kaolin and bentonite were purchased from Baker Chemical Co. Ordi- nary quartz sand was round in a Beuhler mill to make

water used was filtered Minneapolis City tap water con- taining about 40 ppm Ca and 15 ppm Mg; it had a pH of 7.4 to 7.8 and a turbidity of about 0.1 nephelometric turbidity unit (NTU). MgO and sand media were also used to filter for an iron ore processing plant water using the 2.5-cm- diameter system. Media depth was 45 cm of 20130 mesh sand or MgO, the flow rate was 100 ml/min, and the alum dosage was 15 ppm.

“Figure 1 shows photomicrographs (x120) of the two forms of MgO; (A) powdered magnesia used for precipita- tion research (vide infra), and (B) granular periclase used for filtration research. Conventional filter sand is more spherical than the periclase shown in Figure 1B.” The per- iclase used in this work was obtained from two sources: Basic, Inc., Cleveland, Ohio, and Kaiser Refractories, Oakland, Calif. Both materials performed equally as filter media, and each was 97 to 98 percent pure. Impurities, in decreasing order of concentration, were the oxides of cal- cium, silicon, iron, and aluminum. The MgO from Basic, Inc., was a “minus 18 mesh” product from which 20/30 mesh and 30/50 mesh fractions were separated by sieving. The Kaiser Refractories material was their “K-grade peri- clase” which is pelletized into nominally 0.5- by l-cm pil- lows. This product was crushed and sieved to give the size fractions used as filter media. Gamet sand and “flint” (sil- ica) sand were purchased from Neptune Microfloc, Inc., Corvallis, Ore., and sieved prior to use.

Metals Precipitation Research

To perform metals removal tests, 250 to 1,000 ml of solu- tion was placed in a beaker. A weighed quantit of pow-

was added all at once, and the H was monitored with a pH

ally. After 20 to 30 minutes reaction time, a 0.05-percent solution of Separan AP-30 (an anionic polyacrylamide) was added, followed by another 5 to 10 minutes of stirring.

fine materia1 from whic a suspensions were prepared. The

ered MgO (Magox HR 98, a product of Basic C B emicals)

meter and recorder as the so P ution was stirred magnetic-

B Figure 1. Photomicrogmphs of powdered MgO (A) ond crushed periclase (B).

Quantities of MgO and Separan AP-30 solution were 0.1 to 1.0 @liter and 1.0 ml/liter, respectively. After flocculation, the suspension was either filtered through Whatman #5 paper or oured into a cone to settle. The filtrate or super-

troscop for residual dissolved metals, and the volume of

used in arallel tests, the procedure was the same, except

9. Separan AP-30 solution was used at 1.0 ml/liter. A so- dium hydroxide solution was tested on two solutions, but since it gave sludge volumes and metal removals similar to lime, it was not studied further. Adding a weighed quan- tity of solid lime all at once gave sludge volumes and met- als removal equal to those using a lime slurry. Using a threefold to fourfold excess of lime gave pH values of 10.5 to 11.5. Sludge volumes and metals removal were the same as at pH 9, except for Mn and Cd, which were more insolu- ble at the higher pH values. MgO and lime were used to treat two ore beneficiation waste waters, one mine drain- age water, and three prepared solutions.

natant so P ution was analyzed by atomic absorption spec-

the sett T ed sludge was measured. When a lime slurry was

that the E ase was added dropwise until the pH reached 8 to

RESULTS AND DISCUSSION Comparison of Filter Media

Parallel tests where MgO was compared with silica sand show the enhanced performance of the former. Results using the 20-cm-diameter filter are summarized in Table 1. During runs of 7 hours with kaolin and fine quartz, MgO allowed about half as much particulate to pass through as did sand; even so, the pressure increase was only 2 KPa compared to about 7 to 10 KPa for sand. Figure 2 shows tur- bidity as a function of time for filtering a 7-ppm quartz SUS- pension using 2.5 p m alum. In filtrations of kaolin SUS-

quartz. Filtrate turbidities for bentonite showed little dif- ference between sand and MgO. However, as with the other particulates tested, pressure increase while using MgO was 2 KPa compared with 11 KPa for the sand. The ability of MgO to retain solids with a lower rate of pressure buildup is due to the loose packing of MgO particles in the filter bed. Sand has about 40% of its bulk volume as voids, while the void volume of MgO is about 55%. Sand of the same nominal grain size would be expected to have a higher straining efficiency, but the positive electrokinetic charge on MgO makes it more effective overall. Figure 3 presents results for iron ore (taconite) processing water in the 2.5-cm-diameter filter, and these follow the results in filtrations of other solids. The filtration rate is 0.20 m/min (5 gpm/ft2). The suspended materials in this water are mainly iron oxide minerals and quartz, but a ain, MgO o -

An on-site test using a 30-cm-diameter pilot system gave results virtually identical with Figure 3.

Table 2 presents results from comparison of 30/5O mesh MgO with 30/50 mesh garnet sand in a 2.5-cm-diameter dual-media filter. The filter consisted of 15 cm of 20130

pensions, filtrate tur K idities were very much like that of

erated at least twice as long as sand before a reakthroug K .

I0.C

3 I- z i t 0 m a 1.0 3 c W c

c J LL

a a -

0. I

KEY A Sand filter medium 0 Magnesium oxide. M9O

i 0 1 2 3 4 5 6 7 8 9 IO

TIME, hr

Figure 2. Filtrate turbidity versus time for the filtration of 7 ppm fine quartz with 2.5 ppm alum through MgO and silica sand ot o flow velocity

of 0.28 m/min.

mesh silica sand followed by 30 cm of 30/50 mesh garnet sand or 30/50 mesh MgO. As with the 20-cm-diameter filter, MgO gave a filtrate of lower turbidity and a lower head loss rate than the garnet sand. For both kaolin and quartz, filtrate turbidities durin the first 2 hours were es-

filtered through garnet sand, the filtrate turbidity began to increase and reached about 1 NTU after 6 hours. In MgO filtration of kaolin, the filtrate turbidity remained very low and increased to only 0.25 NTU in 6 hours. For filtration of the quartz suspension, a breakdown of per- formance of the garnet sand filter occurred after about 3 hours (see Table 2). MgO removed over 97% of the sus- pended solids even after 5 hours. The quartz suspension contained over four times as much solids as the kaolin sus- pension, causing the filters to become satruated and lose effectiveness much more rapidly.

MgO filters were cleaned between filter runs by meth- ods similar to conventional backwashing techniques. Water and air were pum ed upward through the 20-cm-

bed expansion. This was sufficient to restore the filter to

sentially the same for both P Iters. When kaolin was

diameter filter for 1 to 2 K ours to give a 10- to 20-percent

TABLE 1. SUMMARY OF FILTRATION RUNS USING THE 20-CM-DIAMETER FILTER A T A FLOW VELOCITY OF 0.28 M/MIN.

Filtrate Filtrate Initial Initial

Suspended concentration, dosage, duration, Filter medium, turbidity, at 1 hour, end of run, drop, drop, solida PPm PPm hr 20/30 mesh NTU NTU NTU KPa KPa

Solids Alum Run Original turbidity turbidity pressure pressure

~- Kaolin 5.0 5.0 7 Sand 3.1 0.19 0.47 8 15

Quartz 7.0 2.5 7 Sand 4.5 .44 .46 8 17

Bentonite 5.0 5.0 7 Sand 1.3 .32 .50 10 21

Kaolin 5.0 5.0 8 MgO 3.3 .18 .15 6 8

Quartz 7.0 2.5 7 MgO 4.8 .43 .20 5 7

Bentonite 5.0 5.0 7 MgO 1.2 .23 .50 7 9

a Suspended solids were added to filtered Minneapolis tap water.

138 May, 1984 Environmental Progress (Vol. 3, No. 2

I0.W

L

d-

i I

f

I ~

.m"

e Procsrrlng P

lan1 WoIcr

lniliol turbidity 56 NTU

A

lum dosoge 15 ppm

t

Flow role

5 pp

mlfl'

3

P

Dl

I 2

3 4

5 6

7

I TIM

E, h

wrs

Figure 3. Filtrate turbidity for Mg

O and sand filtration of toconite pro-

cessing water.

k

El

Tr 09 its original perform

ance. Because of the sm

all size, the 2.5-cm

filter could not be cleaned by backwashing. T

he

determine the long-term

stability of MgO

fil filter m

edia was used for 50 cycles of filtering a 25-ppm

quartz suspension w

ith only air-water backflushing be-

tween runs. T

he testing was spread over one year, and

filtrate turbidities and head loss rates were the sam

e fo the last runs as for the first ones. N

o increase in MgO

fines w

as noted from repeated backw

ashing. Although nc

Inc

s n

f nprfnrmance or m

edia was observed for filterin!

19

media w

as emptied into a beaker, stirred, and rinsed. To

E

-! El 4

ters, the same

4

e 4

t- 1

z co c?

r re W

J 7

r-----

__ neutral suspensions, M

gO is a basic m

aterial so it would

dissolve if used to filter acidic solutions. U

nbuffered water initially at p

H 6 to 7 has the pH

raised igh M

gO. T

he to about 8.0 to 8.5 w

hen it is filtered throu dissolved M

g+2 concentration generally increases filtration by 1 ppm

or less.

Metal C

oncentrations M

; during

-0

c -0

IB

2

Table 3 presents the pH

and compositions of the w

aters that w

ere treated. It also compares final m

etal concentra- tions w

hen MgO

and lime are used to treat them

. When

equal pH values are attained, M

gO leaves less dissolved

metal and less m

etal hydroxide suspended if the solids are separated by sedim

entation. The m

ore complete sedim

en- tation of the M

gO-m

etal hydroxide recipitate, com

pared w

ith the lime precipitate, is due to t e low

bulk density of the latter.

MgO

can remove any m

etal that is precipitated as the hj d

rnxirle

. The solubilities of the m

etal hydroxides deter-

K 1

W E

1-

I-

-._-I

- - .

mine the required pH

, and it is adjusted by varying the quantity of M

gO. For exam

ple, manganese requires a

higher pH than zinc or copper, w

hich in turn requires a higher value than iron. T

able 3 also shows the final pH

at different M

gO dosages for several process w

ater samples.

Generally, a threefold to fourfold stoichiom

etric excess of M

gO is required to reach adequate pH

values, which are

comm

only 8 to 9. I* 0

Choracteristics o

t Sludge

A principal reason for usin M

gO instead of lim

e is th low

er volume of hydroxide s udge. T

able 4 resents data

showing that the lim

e precipitate occupies a out three t five tim

es as much volum

e. In most cases sludge has to b

dewatered or handled in som

e way, and the m

ore com act

material from

MgO

is clearly advantageous. On stan ing

for about a day or longer, this sludge self-c stable m

aterial. The tw

o cones in the photograph (Figure 4) had sedim

ent collected in the bottom, and the conc

were tipped to a horizontal position after about 24 hours s

the water poured out. T

he photographs were then taken

from above. T

he sludge on the left was precipitated w

ith

K P

- ,

P

1

e 0

.d

:0

le

cements into a

?S

io

1 TABLE 3. RESULTS FROM TREATING WATER WITH MGO AND LIME 8”

Beneficiation process waste-CM;

0.1 dlitelg MgO, filtered 0.2 dliter MgO, filtered 0.35 dliter MgO, filtered

Beneficiation process waste-BM;

Untreated water Treated with:

Untreated water Treated with:

0.16 @liter MgO, filtered 0.21 ,@iter MgO, filtered 0.31 ejliter MgO, filtered

Mine drainage; Untreated water Treated with 0.5 #liter MgO,

filtered Prepared solution # 1;

Untreated water Treated with:

0.4 dliter MgO, filtered 0.4 dliter MgO, settled 0.1 dliter lime: settled

Prepared solution #2; Untreated water Treated with:

0.125 g/liter MgO, filtered 0.04 @liter lime: filtered

Prepared solution #3; Untreated water Treated with:

0.2 g/liter Mg, settled 0.05 @iter lime,” settled

PH -

5.4

8.6 9.2 9.4

6.4

8.3 8.7 8.9

2.7 8.9

4.2

8.9 8.9 9.4

5.4

8.9 8.9

4.0

9.0 9.0

I 8

Cd Pb ’ d.

4% Chemical Analysis, ppm -

Fe -

5.7

<0.2 < .2 < .2

< .2

ND ND ND

40 .2

ND

N D ND ND

5.0

< .2 < .2

N D

N D ND

c u

0.63

0.1 .1

< .1

< .1

ND ND ND

ND ND

8.7

< .1 .5 .7

.21

< .1 < .1

N D

N D ND

Zn

0.55

<0.1 < .1 < .1

12.7

< .2 < .2 < .2

39 .1

N D

N D N D N D

2.7

< .1 .2

4.2

< .1 .8

Ni -

NDb

ND N D N D

<0.2

N D N D ND

ND N D

12.0

< .2 .2

1.2

N D

ND ND

ND

N D ND

‘The unit dliter is the grams of MgO used per liter of water treated. ND = not determined, since initial concentrations were below the analysis limit of atomic absorption Weight o f lime is for CaO, not Ca(OH)l.

MgO, while that on the right was not; the greater stability of the MgO sludge is obvious.

Reoction Times and MgO Dosages

Figure 5 shows a plot of pH vs. reaction time for a range of MgO dosages used to treat process water BM. This water is a mixture of tailings pond overflow and mine water from a Western lead-silver operation. The greater the quantity of MgO used, the more rapidly the pH reached its maxi- mum value. For example, addition of 0.35 g/liter gave es- sentially the final pH in 5 minutes, but at 0.21 g/liter, the pH is still increasing slowly after 12 minutes. For applying MgO to a full-scale process, a compromise would have to be made between minimum chemical usage on the one hand and sufficiently high pH and minimum reaction time on the other.

For an ideal reaction, the required amount of solid MgO must be added all at once. This permits the slow solid- liquid reaction that is required for a compact precipitate. In one experiment, MgO and water containing metals were continuously added to a stirred beaker. The pH re- mained at about 9, so the metals reacted with hydroxide in solution rather than with solid MgO. The overflow was collected and allowed to settle, but the precipitate resem- bled the lime precipitate since it was voluminous and did not cement on standing. In another experiment, a 10-percent slurry of MgO was made and applied in the proper amounts to precipitate dissolved metals. When this slurry was used immediately, results were the same as with solid MgO. However, if the slurry had been prepared for more than 1 to 2 hours before use, the increase in pH was much slower. The sludge was compact, but it did not self-cement,

Mn -

9.9

7.1 3.4 1.3

17.5

8.3 5.7 1.9

41 15

ND

N D ND ND

4.4

< .2 2.2

ND

N D ND

C O -

N D

N D N D N D

N D

N D N D N D

N D N D

11.1

< .2 .2

1.6

N D

ND ND

ND

ND ND

- N D

N D N D N D

N D

N D N D N D

N D N D

N D

N D N D N D

ND

ND N D

5.2

.31 1.4

- ND

ND ND ND

ND

ND ND ND

ND ND

ND

ND ND ND

ND

ND ND

4.7

< .5 1.6

Disadvantages of MgO in Metals Precipitation

The major disadvantage of MgO is that it costs more than some other basic materials. However, when the total met- als content is low, chemical reagents account for only a small fraction of the treatment cost. The savings from easier sludge dewatering, compactness, and stability may more than make up for added reagent costs. If the metal concentration is below 25 to 50 m liter, MgO usage will be about 0.25 g/liter(2 pounds per t f ousand gallons). MgO costing $330/ton used at this rate would result in chemical expenses of about $0.08 per m3($0.31 per 1,000 gallons) of water treated. Lime would cost about one eighth as much,

Operation of a continuous rocess based on MgO is not as straightforward as one w hp ere lime or NaOH is used. Since the metals must react directly with solid MgO, a plug-flow or parallel batch reactor system is required to give continuous operation and ensure a compact sludge. Solid reagent is generally not so convenient as a slurry or solution.

TABLE 4 . SLUDGE VOLUMES F O R T R E A T M E N T OF PROCESS WATERS WITW MGO AND LIME

Sludge volume, %” Lime, 0.14 g/liter MgO, 0.55 @iter

Abandoned metal 2.4 0.6 mine drainage

Lime, 0.025 @liter MgO, 0.1 @liter

Beneficiation process 0.9 0.2 waste-BM

a Volume of sludge per 100 volumes of untreated water.

Environmental Progress (Vol. 3, No. 140 May, 1984

F i 4 . Photogmph showing cemented MgO sludge (left cone) compared with norm1 lime sludge after slight agitation of the sludge.

coNcLusloNs The comparative tests described in this paper show

MgO (periciase) to be superior to silica sand orgarnet sand for filtration of several different particulates. Head loss occurs at a lower rate with MgO, tirbidities are lower, and filter runs can be longer. MgO could be used in sand in existing equipment, and regeneration o f t e filter can be accomplished by conventional backflushing.

Powdered MgO (magnesia) removes heavy metals from water to give a compact sludge that self-cements on stand- ing. MgO precipitates metals as well or better than lime, and the MgO-hydroxide sludge is more easily filtered or dewatered. Several minutes reaction time is required for solid MgO. The cost of MgO is greater than the cost of lime but, for water containing low concentrations of dissolved metals, the savings in sludge handling may more than com- pensate for the cost difference. Future work should deter- mine whether MgO can conveniently be used on a contin- uous basis for heavy metal removal.

RlaCe Of

UTERATURE CITED 1. Weber, W. J., “Physicochemical Processes for Water Quality

Control,” Wiley-Interscience, New York (1972), Chaps. 2 and 4.

2. Kirk-Othmer, “Encyclopedia of Chemical Technology,” Wiley-Interscience, New York(1980), Vol. 10, pp. 284-337 and 489523.

9.0 n ” V “ 8

I 0 1.5 3.0 4.5 6.0 7.5 9.0 12.0 13.5

TIME. mln

7.0

Figure 5. pH-time curves for the treatment of process water BM with various amounts of MgO.

3. Parks, G. A., “The Isoelectric Points of Solid Oxides, Solid Hy- droxides. and Aaueous Hydroxo Complex Systems,” Chem. Reoiews,’ 65, 175208 (1965).

4. Oulman, C. S.. D. E. Burns, and E. R. Baumann, “Effect on Fil- tration of Polyelectrolyte Coatings of Diatomite Filter Media,” Jour. AWWA, 56, 1233-1239 (1964).

5. Baumann, E. R. and C. S. Oulman, “Polyelectrolyte Coatings for Filter Media,” Proc. Filtrution SOC., 682-690 (1970).

6. Knight, R. A. and P. Marinaccio, “Process for Producing MicroDorous Films and Products.” U.S. Patent 3,876,738 . .

(AprilL8, 1975). 7. Lanouette, Kenneth H., “Heavy Metals Removal, Chem. Eng.

84. NO. 22, 73-80. 8. Jorgensen, S. E., “Industrial Waste Management,” Elsevier

Scientific Publishing Company, New York, N.Y. (1979), pp. 187-214.

Joseph E. Schiller received his undergraduate training at Central Michigan University and a Mas- ters degree in Chemistry in 1968 from Michigan State University. Following two years ofjuniorcol- lege teaching, he returned to graduate school at the University of North Dakota when he received a Ph.D. in 1973. He is presently employed as Direc- tor of Refining Operations at Materials Processing Corporation, St. Paul, Minn.

Daniel N. Tallman received his B.A. in Chemistry in 1980 from Augsburg College, Minneapolis, Minn. He is a research chemist with the Twin Cit- ies Research Center of the US. Bureau of Mines. His research interests involve treatment ofmineral process waters, filtration, removal and detoxifica- tion of cyanides, and removal of heavy metals from mine waste streams. He is a member of the Ameri- can Chemical Society.

Sanaa E. Khalafalla received his Ph.D. in Physical Chemistry from the University of Minnesota in 1953. His current research includes chemical and electrochemical aspects of drilling, dust physical chemistry and wettability, mineral process waters and spontaneous combustion. He is a member of AIME and past chairman of the Physical Chemis- try Committee of TMS-AIME. He has authored over 120 publications and holds 15 U.S. Patents.